Alkali-treated titanium dioxide promotes formation of proteoglycan layer and altered calcification and immunotolerance capacity in bone marrow stem cell

Introduction In this study, we report that a proteoglycans (PGs)-layer between the bone and titanium dioxide (TiO2) surface after osseointegration improved the calcification capacity and immunotolerance of human bone marrow mesenchymal stem cells (hBMSCs) on TiO2. Alkaline treatment of TiO2 is a method for promoting osteogenesis in hBMSCs. We hypothesized that promotion of osteogenesis due to alkaline treatment was caused by changing PGs-layer on TiO2. Objective This study aimed to analyze whether alkaline treatment of TiO2 affects PGs-layer formation and immunotolerance in hBMSCs. Methods The topology and wettability of the alkaline-treated titanium (Ti-Al) and unprocessed titanium (Ti-MS) surfaces were characterized. Initial cell attachment, cell proliferation, calcification capacity, alkaline phosphatase activity, PGs-layer formation, PGs function, and the expression of osteogenic and immunotolerance-related genes were analyzed. The conditioned medium (CM) from hBMSCs grown on Ti-Al and Ti-MS was added to macrophages (hMps) and Jurkat cells, and immunotolerance gene expression in these cells was analyzed. Results hBMSCs cultured on Ti-Al showed increased initial cell attachment, cell proliferation, PG-layer formation, and osteogenic capacity compared with hBMSCs on Ti-MS. Gene expression of indoleamine 2,3-dioxygenase (IDO) in the hBMSCs cultured on Ti-Al was higher than that in the hBMSCs on Ti-MS. CM from hBMSCs did not affect markers of M1 and M2 macrophages in hMps. CM from hBMSCs cultured on Ti-Al altered the gene expression of Foxp3 in Jurkat cells compared to that of CM from hBMSCs on Ti-MS. Significance These results suggest that alkaline treatment of TiO2 altered PGs-layer formation, and changed the osteogenesis and immunotolerance of hBMSCs.

A dental implant is an artificial root used in dentistry to support a prosthetic restoration with the goal of replacing missing teeth.Dental implant treatment success requires osseointegration, which involves direct contact between the bone and dental implant.This phenomenon occurs when BMSCs adhere and differentiate into osteoblasts on the implant surface.Titanium dioxide (TiO 2 ) has been widely utilized for dental implants due to its positive influence on the wound healing process by virtue of its biocompatibility and mechanical strength [6], in addition to being an abundant natural resource.Various modifications to TiO 2 surface have been tried to improve osseointegration [2,7], because the rate of success is about 95% [8] but not 100%.
Alkaline treatment of the TiO 2 surface is simple and economical, and the nanotube dimensions can be precisely controlled compared with other treatments to improve osseointegration.Alkaline treatment changes the topology of the TiO 2 surface into a micro/nanostructure and improves its wettability [9], which improve the attachment, proliferation, and osteogenic differentiation in BMSCs [1,[10][11][12][13][14][15][16][17][18], even bone-induction capacity in vivo and in vitro [2,[19][20][21].Alkali-treated TiO 2 affects its biocompatibility by controlling macrophage differentiation into the M2 phenotype, which modulates the inflammatory response [22][23][24].However, the mechanisms of osteogenic capacity and biocompatibility of alkali-treated TiO 2 and the molecular interactions that occur at the implant interface are yet to be fully elucidated.
Proteoglycans (PGs) and glycosaminoglycans (GAGs) are molecules abundant on the cell surface and in the extracellular matrix where they play pivotal roles in facilitating signal transduction and maintaining stem cell homeostasis [25], and also serve as key regulators of immune cell (leucocyte) recruitment and positioning during the inflammatory process [26].PGs modulate osteogenic differentiation of MSCs [27].50-500 nm of PGs was observed between the bone and TiO 2 surface, and this layer contained PGs without collagen fibers as the extracellular matrix (ECM) after osseointegration, as observed by transmission electron microscopy (TEM) [28].PGs are heavily glycosylated by chondroitin sulfate (CS), dermatan sulfate (DS), keratin sulfate, and hyaluronic acid.These polysaccharides, consisting of repeating disaccharide units, are called glycosaminoglycans (GAGs), and are catalyzed by glycosyltransferases. PGs and GAGs are abundant on the cell surface and in the extracellular matrix where they play pivotal roles in facilitating signal transduction and maintaining stem cell homeostasis [29].PGs modulate MSC osteogenic differentiation of mesenchymal stem cells [30].The sulfation contained in PGs affects the osteogenic differentiation of stem cells because sulfate groups provide binding sites for multiple growth factors [31].
A previous report suggested that chondroitin-4-sulfate (C4S), a GAG, adsorbed on TiO 2 strongly, compared to other glycans [32].The role of the PG-layer on TiO 2 is biological interfacial adhesion between the bone and titanium surface [33], but it does not alter the intrinsic strength of the mineralized tissue [34].C4S transferase-1 (C4ST-1) catalyzes 4-O-sulfation in CS [35] and the PG-layer between hBMSCs and TiO 2 is altered by C4ST-1, directly and indirectly affecting the immunotolerance capacity but not the osteogenic properties of hBMSCs on TiO 2 [36].Thus, the role of the PG-layer on TiO 2 has not yet been fully elucidated.
We hypothesized that the alkali treatment of TiO 2 affects the morphology and function of the PG-layer derived from hBMSCs.To test this hypothesis, we analyzed the gene expression of glycosyltransferases and the calcification capacity of PGs in BMSCs cultured on alkali-treated TiO 2 in vitro.Furthermore, we analyzed the immunotolerance of BMSCs, T-cell, and human macrophages (hMps) on alkali treated-TiO 2 .

Specimen preparation
Ti disks (80, 32, 15, and 6 mm in diameter and 2 mm in thickness) of grade 2 commercially pure Ti were prepared by machining (Ofa Co., LTD, Chiba, Japan).After polishing with #800 grinder, Ti O 2 disks were coated with 4.0 mM HNO 3 and 0.8 mM HF, then washed ultrasonically in 1% (w/v) sodium dodecyl sulfate (SDS) (FUJIFILM-WAKO, Osaka, Japan), acetone (FUJIFILM-WAKO), and 70% ethanol, for subsequent use.These TiO 2 disks were immersed and treated with an ultrasonicator in 5.0 M NaOH aqueous solution and placed in bath maintained at 37 • C for 1 h (Ti-Al).Unprocessed TiO 2 disks were used as controls (Ti-MS).

Surface characterization
The topographies of Ti-MS and Ti-Al were observed using highresolution scanning electron microscopy (SEM) (S-4800; Hitachi, Tokyo, Japan).Wavelength dispersive spectrometry (WDS) was performed to evaluate the chemical composition of the sample surface.Surface wettability was evaluated using a contact angle measurement system (VSA 2500 XE; AST Products, Billerica, MA, USA).

Cell proliferation assay
The 6 mm Ti disks were placed in 96-well plates.After seeding 1 × 10 3 cells on Ti-MS and Ti-Al disks for 1, 3, 7, 10, and 14 days, proliferation assays were carried out using the Cell Counting Kit-8 (Dojindo Laboratories, Kumamoto, Japan), in accordance with the manufacturer's instructions.Absorbance was determined using SpectraMax 340PC384 (Molecular Devices, Tokyo, Japan).

Measurement of alkaline phosphatase (ALPase) activity
UE7T-13 cells were plated at a density of 5 × 10 3 cells on the samples of 15 mm diameter of Ti-MS and Ti-Al.For quantitative analysis of ALPase activity, p-nitrophenol production was measured at 37 • C for 6 min in Milli-Q water using a SIGMAFAST p-Nitrophenyl phosphate Tablet set (Sigma-Aldrich, St. Louis) as a substrate.The relative amount of p-nitrophenol was estimated from the light absorbance at 405 nm on days 7, and 14 using the SpectraMax 340PC384 (Molecular devices).

Alizarin red S staining (ARS)
For staining, 2 × 10 4 cells were seeded on Ti-MS and Ti-Al (15 mm in diameter) and cultured with DMEM supplemented with 50 μg/mL ascorbic acid (FUJIFILM-WAKO), 100 nM dexamethasone (FUJIFILM-WAKO), and 10 mM β-glycerophosphate (FUJIFILM-WAKO) for 7 and 14 days.Each sample was fixed with 10% formalin and stained with mM ARS solution (FUJIFILM-WAKO) for 30 min at room temperature.After staining, the cells were washed five times with distilled water.The dye was then dissolved using 500 μL of 5% formic acid for 20 min.The extract from 2 M NaCl were diluted into 1 μg/μL with PBS and seeded μL on a 10 mm square Ti-MS and Ti-Al.After incubation, samples were rinsed three times with PBS, and 100 μL of 1 mM CaCl 2 was added for another overnight incubation at 37 • C 40 mM ARS solution was added to samples for 30 min at 37 • C. the semi-quantification of ARS was as mentioned above.

Immunocytochemical (ICC) analysis
Cells were seeded on Ti disks at a density of 0.5 × 10 2 /cm 2 and cultured for 24 h.Cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100 in PBS for 10 min.Each disk was stained with DAPI (Vector Laboratories, Inc., Burlingame, CA, USA) and observed under an FV3000 Confocal Laser Scanning Microscope (Olympus, Tokyo, Japan).
T. Mizutani et al.

TEM (transmission electron microscopy)
5 × 10 2 /cm 2 cells were seeded on the Ti foil discs of 6 mm diameter.After 7 and 14 days cultivation, samples were washed three times with 0.1 M sodium cacodylate buffer (Nisshin EM Co., Ltd., Tokyo, Japan), fixed with a pre-fixative solution (4% paraformaldehyde (Nisshin EM Co., Ltd), 2% glutaraldehyde (FUJIFILM WAKO), and 0.1 M sodium cacodylate) for 1 h at 4 • C, and immersed in 0.05% ruthenium red solution for 1 h at room temperature A post-fixation was performed with 1% OsO 4 for 1 h at room temperature, followed by dehydration with ethanol and embedding in EPON resin (Nisshin EM Co., Ltd).Samples were cut with an ultramicrotome (Ultra CUTS, Leica, Tokyo, Japan), stained with lead citrate for 5 s, and observed with a transmission electron microscope (JEM-1400PLUS, JEOL).

Real-time RT-PCR
Cells were seeded on the 32 mm diameter of Ti-MS and Ti-Al at a density of 2 × 10 4 /cm 2 and cultured for seven and 14 days.Total RNA was isolated using TRIzol LS Reagent (Thermo Fisher Scientific, Waltham, MA, USA), and cDNA was synthesized using ReverTra Ace qPCR RT Master Mix and gDNA Remover (Toyobo Co., Ltd., Osaka, Japan) according to the manufacturer's protocols.Quantitative real-time RT-PCR was performed using THUNDERBIRD SYBR qPCR Mix (Toyobo) and the Mx3000P Real-Time QPCR System (Agilent Technologies International Japan, Ltd., Tokyo, Japan).The expression of target transcripts was normalized against the expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and relative changes in gene expression were determined using the 2 − ΔΔCt method.

Immunotolerance assay of hBMSCs, hMps, and T-cells
UE7T-13 cells were cultured on 32 mm Ti-MS and Ti-Al disks for and 14 days.The medium was replaced with DMEM without FBS for h to prepare the conditioned medium (CM).CM was then added to hMps and Jurkat cells and cultured for 24 h.Total RNA was extracted to analyze M1 and M2 macrophage polarization and regulatory T cell (Treg) markers.

Image and statistical analyses
One-way analysis of variance followed by the Student's-Newman-Keuls post-hoc test was used to determine significance.Statistical significance was set at P < 0.05.

Surface characterization of Ti-MS and Ti-Al
Ti-MS surfaces showed typical machined surface topology (Fig. 1A, left panel).The Ti-Al surface exhibited an alveolate porous structure with homogeneous holes (Fig. 1A, right panel).The results of the elemental analysis of the material surface are shown in Fig. 1B.The atomic and mass percentages of Ti on the surface of Ti-MS were significantly lower than those of Ti-Al.The amounts of C and O on the Ti-Al surface were significantly higher than those on Ti-MS.No other elements were detected in Ti-MS and Ti-Al.Hydrophilicity analysis showed that the Ti-Al surface exhibited more hydrophilic properties than the Ti-MS surface (Fig. 1C).

Cell attachment and proliferation on Ti-MS and Ti-Al
Initial cell adhesion of UE7T-13 on Ti-Al was significantly higher than that of UE7T-13 on Ti-MS (Fig. 2A).At all time points, the number of cells on Ti-Al was significantly higher than that on Ti-MS (Fig. 2B).

Osteogenic activity of UE7T-13 on Ti-MS and Ti-Al
Runt-related transcription factor-2 (Runx2), is known as a regulator of bone development.Runx2 controls osteocalcin (OCN) and alkaline phosphatase (ALP) expression which play important roles in mineralization.Type I collagen (Col-I) gene and its product is the main organic component in bone and serves as a template for mineralization [38].These genes are markers of BMSC differentiation from osteoprogenitors to osteoblasts.Runx2, OCN, and ALP expression in UE7T-13 on Ti-Al was significantly higher than that on Ti-MS after two weeks of cultivation; however, no significant differences were observed after one week of cultivation.Furthermore, there were no significant differences in Col-I expression between the UE7T-13 cells cultured on Ti-MS and Ti-Al (Fig. 3A).The ALPase activity in UE7T-13 on Ti-Al was significantly higher than that in UE7T-13 on Ti-MS at 1 and 2 weeks (Fig. 3B).The number of calcium nodules stained with ARS in UE7T-13 on Ti-Al was significantly higher than that in UE7T-13 on Ti-MS after 2 weeks of cultivation (Fig. 3C).

Expression of glycosyltransferase in hBMSCs on Ti-MS and Ti-Al
After 1 week of cultivation, the expression of XylT-1 in UE7T-13 on Ti-Al was significantly lower than that in UE7T-13 on Ti-MS.After two weeks of cultivation, the expression of all glycosyltransferases in UE7T-13 on Ti-Al was significantly higher than that in UE7T-13 on Ti-MS (Fig. 4).

Ultrastructural analysis of the PG-layer
At 1 and 2 weeks, PGs formed on Ti-Al had needle-like structures extending across the UE7T-13 interface.However, the PGs formed on Ti-MS did not exhibit a spike-like shape or vague staining between UE7T-13 and Ti-MS (Fig. 5A).The area and relative gray value of the PG-layer formed on Ti-Al were significantly higher than those formed on Ti-MS after 2 weeks of cultivation (Fig. 5B).

Immunotolerance capacity of UE7T-13 on Ti-MS and Ti-Al
The gene expressions of PGES2, PD-L1, and AHR were not significantly different in UE7T-13 between Ti-MS and Ti-Al at any time point.The expressions of IL-6 and IL-10 in UE7T-13 on Ti-Al were significantly lower than those in UE7T-13 on Ti-MS after one week of cultivation; however, the difference was not significant after two weeks of cultivation.The gene expression of IDO in UE7T-13 cells on Ti-Al was significantly higher than that in UE7T-13 cells on Ti-MS at all time points (Fig. 6).

Immunotolerance capacity of hMps and T-cell affected by UE7T-13 on Ti-MS and Ti-Al
Gene expressions of CD86, CD163, CD206, and IDO in hMps treated with CM derived from UE7T-13 cultured on Ti-Al were significantly different from those in hMps treated with CM derived from UE7T-13

Table 1
Weight % and Atomic % of components on Ti-MS and Ti-Al disks by SEM-WDS.

Discussion
In this study, we analyzed whether alkali-treated TiO 2 affected the immunotolerance and osteogenic capacity in BMSCs.We first characterized the surface of Ti-MS and Ti-Al, and the results were similar to those found in a previous study [9].These results suggested that treatment of ultrasonication was useful, because the treatment time could be shortened.
The results of the initial cell attachment and cell proliferation also corresponded to those of previous studies.In contrast, the expression of osteogenesis-related genes exhibited some differences [18,19].In previous reports, the expression of Col-I in BMSCs cultured on Ti-Al was significantly higher than that in BMSCs cultured on Ti-MS.This difference was suggested to be caused by differences in the cells.In contrast, the calcification capacity of the hBMSCs on Ti-Al was consistent with previous reports.The alkali treatment of TiO 2 in this study promoted osteogenic properties compared with non-treated TiO 2 .
We investigated the relationship among the expression of glycosyltransferases, calcification capacity, and initial cell attachment in hBMSCs cultured on Ti-MS and Ti-Al.FAM20B is a glycan kinase essential for the morphogenesis and mineralization of the craniofacial complex [39,40].Increased levels of XylT-1 correlate with mineralization of the extracellular matrix during osteogenic differentiation of mesenchymal stem cells [41].Highly sulfated CS promotes osteogenic differentiation because the 4,6-disulfates in N-acetyl-galactosamine provide binding sites for these protein ligands [30,42].Enzymatic digestion of CS impairs the formation of mineral modules [43].CS and DS play important roles in cells involved in osteoblast-like cell adhesion [44].The results of the present study are consistent with those of previous studies.FAM20B, XylT-1, and CS have been suggested to play important roles in calcification and cell proliferation on TiO 2 .
Ultrastructural images suggested that a PG-layer formed along the fine protrusions of Ti-MS and Ti-Al as a template, and surface roughness on Ti-Al was suggested to contribute to the promotion of the area of the PG-layer.Needle-like crystals composed of mineral substances and/or calcified nodules have been observed in collagen fibers in bone tissue [45,46].Furthermore, these results correspond to those identified using ARS.It has been suggested that the calcified nodules on Ti-Al are more similar to those on natural bone.Thus, the surface modification of TiO 2 changed the morphology, area, and function of the PG-layer.
The chromatograms of PGs eluted from hBMSCs on Ti-MS and Ti-Al corresponded to a previous study and were shown to extract proteoglycan in a 2.0 M NaCl fraction [47].However, these results did not correspond to ultrastructural observations because absorbance at 280 nm only detected amino acids and not glycan.The smeared bands of PGs in hBMSCs cultured on Ti-Al did not shift the low molecular weight in Stains-All staining.This suggests that the effects of the surface topology and wettability of Ti-Al on GAG quantity in the PG-layer were relatively small.Furthermore, the calcification capacity results suggested that PGs derived from BMSCs cultured on Ti-Al had the ability to improve mineralization compared to those cultured on Ti-MS.Previous reports have shown that the modification of GAG components in biomaterials changes osteogenic activity in the field of bone regeneration [48].However, the detailed components of the GAGs containing PGs on Ti-MS and Ti-Al were not revealed in this study.Future studies should investigate the GAG components of individual PGs, such as decorin and biglycan.
During the healing phase of tissue repair and immune-mediated diseases, mesenchymal stem cells, including BMSCs, control immune cells such as T cells, macrophages, and dendritic cells [49,50].We previously reported that C4S, a GAG, affects the immunotolerance of BMSCs on TiO 2 [3].Alkali treatment of TiO 2 promoted the polarization of macrophage-like cell lines into M2 macrophages [22,51,52].Our results suggest that BMSCs on Ti-Al directly promote immunotolerance.BMSC-derived IL-6 and PGE2 stimulate M2 macrophage differentiation and IL-10 secretion from macrophages, thus reducing inflammation [53,54].Moreover, BMSC-derived IL-10 generated M2 macrophages, which create an immunosuppressive microenvironment [49].In our results, hMps tended to differentiate into M2 macrophages, but did not differentiate completely because CD80 expression was not significantly different between the two experimental groups.Our results suggest that the difference in IL-6 and IL-10 expression between the two experimental groups was too small to differentiate macrophage polarization, because differences in gene expression were approximately 2-to 3-fold.IDO is the rate-limiting enzyme in the tryptophan catabolic pathway and is regarded as one of the key modulators of acquired immune tolerance [55].These results suggest that the PG-layer controls immunotolerance around TiO 2 directly and indirectly.Furthermore, IDO affects the generation and expansion of Tregs [56].Tregs are identified by the expression of CD4, CD25, and Foxp3 [57,58].However, the expression of the Treg marker in Jurkat cells treated with CM derived from hBMSCs cultured on Ti-Al was atypical.CD25 and Foxp3 expression was very low because the cycle threshold of CD25 and Foxp3 in Jurkat cells ranged from 30 to 33 (data not shown).The cycle threshold of Foxp3 in Jurkat cells treated with DMEM without FBS was in a similar range, 31-33 (Supplemental data.1).On the other hand, Foxp3 expression in Jurkat cells is originally higher than that in other cancer cells [59].From the above, it was suggested that low expression of Foxp3 in this study was caused by the lack of FBS in CM.Stronger stimulation, such as co-culture system [60], gene transduction [61], IDO inhibitor [62], and suppression assays [63] will be needed for Jurkat cells to differentiate into Tregs.In the future, we should investigate the proteins and micro-RNA in CM to determine the factors that induce the differentiation of Jurkat cells into Tregs.Furthermore, we will investigate the function of identified factors in BSMCs on Ti-MS and Ti-Al to immune cells using gene silencing methods, such as small hairpin RNA.

Conclusion
This study revealed that alkali-treated TiO 2 promotes the gene expression of glycosyltransferase in hBMSCs and alters the morphology and quantity of the PG-layer that was formed by hBMSCs.Alkali -treatment of TiO 2 promoted the calcification capacity of PGs derived from hBMSCs and affected their immunotolerance capacity of hBMSCs both directly and indirectly in vitro.

Fig. 1 .
Fig. 1. (A) SEM micrograph images of surface of Ti-MS and Ti-Al.(B) Characterization of TiO 2 surface using WDS analysis.WDS-spectrum quantification of individual elements is shown in Tables 1 and 2. (C) The contact angle on the surface of Ti-MS and Ti-Al.Notes: Data are expressed as means (n = 6) with error bars representing standard deviations; *p < 0.05.

Table 2
Elemental composition of samples.